Wednesday, November 02, 2011
To this group, the Internet is a must-have. To some, particularly those located in India, China, Germany and Japan, Internet access is more important than owning transportation. A little over one-third (32%) of College Students indicate the Internet is as important to them as water, food, air and shelter. 40% of all college students consider Internet access more important than dating and social activities. Over 50% of both groups said they could not live without Internet access.
Seriously?? Can't live without it?
Okay, aside from the over-the-top answers like, "can't live without it", it's worth it to download the reports and review the information. I have yet to get through it all but there were a few things that caught my eye.
Being someone that follows the chip industry I will admit to cracking a smile when I saw that 45% of the College Students and 36% of the Employees said laptop computers are the primary tool they use to get information and news. As much as we like to talk about mobile phones and the growth they are enjoying, in China 2/3rds, or 67%, of their College Students were using laptops to get their Internet content.
(Hey! I thought the PC was dead?)
Smartphones are important. They were as used as the primary device for finding information and news by 10% of the College Students and 11% of the Employees. In Spain, 36% of the College Students were using a smartphone as the primary way to get information. In Australia, 1 in 4 Employees were using a a smartphone as their primary gathering tool.
The type of hardware used tends to vary quite a bit when looked at on a regional basis but one over-riding theme is very clear, no matter what the device is, College Students and Employees would like to be able to choose what they use to access networks anywhere, anytime.
If you are interested in these trends there's much more in the report.
Monday, October 31, 2011
Thailand floods take toll on PC makers.
A quote from iSupply:
“Beyond Thailand itself, the worst-impacted country is Japan, which maintains extensive manufacturing operations in areas affected by the disaster,” said iSuppli in a statement.
“Overall, the Thailand flooding represents the second major natural disaster to affect Japan this year, after the March earthquake. Thailand plays a key role in the manufacturing operations of Japanese companies, with an estimated 1,800 Japanese manufacturers operating in that country and 450 Japanese businesses located in seven flood-hit industrial parks,” the company added.
Here are some photos of what has happened in Japan six months after the tsunami:
Japan Marks Six Months Since Earthquake, Tsunami
You have to hand it to them for their resilience.
Second Energy Department-backed company goes bankrupt - The Hill's E2-Wire.
Beacon Power Corp., which develops energy storage systems, filed for bankruptcy protection in the U.S. Bankruptcy Court in Delaware.
Beacon Power had received a federal loan guarantee to help build an energy storage plant in Stephentown, N.Y., that began operating in January. The Treasury Department’s Federal Financing Bank provided the loan.
Beacon sought bankruptcy protection two days after the White House ordered an independent 60-day evaluation of the Energy Department's loan programs aimed at ensuring effective management and monitoring.
The review, conducted by a former Treasury Department official, will include examination of how Beacon’s project is performing going forward, and whether there are additional steps that can be taken to protect taxpayers, according to the Obama administration.
Beacon is a publicly traded company. At one time, early in this century, the shares fetched over $90. The company has never made any money. Grab a stiff drink and check out the Beacon website if you would like to run through their annual financial performance. ahem
Business Week and Bloomberg have even more on the story:
Beacon Power, Backed by US Guarantee, Files for Bankruptcy
Beacon plunged 33 cents, or 73 percent, to 12 cents at 2:23 p.m. New York time in Nasdaq trading, after dropping as much as 76 percent. Before today, the shares had declined 80 percent this year, prompting Nasdaq to warn Beacon that its stock might be delisted because the price had fallen below the $1 minimum.
The stock closed today's trading session at $0.11 per share - well on its way to doughnut land.
This is far from the end. I fear we will be hearing about a lot of troubles with these alt+energy loans.
Kauffman Economic Outlook: A Quarterly Survey of Leading Economic Bloggers, Third Quarter 2011.
Optimism is out; pessimism is in among the country's top economics bloggers as they look to 2012 and beyond, particularly regarding jobs. A new Ewing Marion Kauffman Foundation survey released today shows that only 50 percent of respondents anticipate employment growth, a decrease of 20 percent from second quarter.
Fully 95 percent of respondents view current economic conditions as "mixed" or "facing recession," an increase of 10 percent from second quarter, and a third predict a double-dip recession during 2012. "Uncertain" is once again the top adjective economics bloggers use to describe the economy, and respondents shared expectations of higher annual deficits and the top marginal tax rate.
From Slide 11 of the Full Survey:
In every category of business, conditions right now are rated as “fair, bad, or very bad” by more than 69 percent of respondents, with venture and angel capital being the least bad of them all at only 69 percent, while bank lending to individuals is rated 90 percent negative.
Well, one thing is for sure, they don't call it the "dismal science" for nothing.
Wait a second - perhaps stuff like this is still a good contrarian play? After all, the stock market, as measured by the Wilshire 5000, just had its biggest up month in 24 years.
Friday, October 28, 2011
Inside WD's flooded Thai factory • Channel Register.
I sent this to a few of my friends in the industry. One industry exec had this to say:
It turns out that Bang Pa in facility was submerged with 2 meters of water. This is the old Read- Write facility and I have been in it many times. The first floor is where most of the slider processing equipment was located. Most of the equipment was lost to the flood. They last saws, grinders, sorters, cleaners, steppers, alignment tool etc. It's a real mess. This situation for WD is pretty bad because they had all their slider fabrication in one basket. It will take at least 4 months to start the recovery and a year to get back to where they were pre-flood. It looks like the high water will not recede for at least 3 weeks.
The issue that is affecting all the drive companies has to do with the supply chain. Hutchinson, Magnecomp, Nidec, On Semiconductor, MicroSemi. All facilities are in similar industrial parks that have been flooded or closed by the flood.
Two Fabrinet facilities are down hard. These two facilities make a large percentatage of optical devices for JDS Uniphase, OCLARO, Finisar and Emcore. On Semi's fab for power and signal management, logic, discrete and custom devices has also taken a hit.
In a follow-up conversation we talked about the equipment and how it would take a long time to replace - perhaps as long as 6 months. One of the reasons for this is that low capital investment rates in disk drive manufacturing pushed many of the companies supplying tools to the manufacturers to down-size or, worse yet, leave the industry. It's not going to be easy for WD to ramp back up.
This story is bound to get more coverage in the coming weeks because it is going to have a significant impact on a host of companies
Tech Industry Urges Congressional Supercommittee to Seize Opportunity to Spur Innovation as Key
Six of America's leading technology trade organizations and their members urged the bipartisan Joint Committee on Deficit Reduction to embrace a series of proposals that would help America's economy and debt stabilization by supporting innovation. The recommendations were part of a letter sent to the Joint Committee today sponsored by TechNet, Business Software Alliance, Consumer Electronics Association, Information Technology Industry Council, Silicon Valley Leadership Group and TechAmerica. The letter included recommendations on tax reform, research and development, spectrum, smart deployment of information technology to reduce waste, high-skilled immigration reform, deployment, among others.
The letter was signed by 50 prominent members of the tech community. Here's the gist of their message:
We write as American business leaders concerned by the direction, substance, and tone of our nation's policies and politics. We are hopeful that your work will help the nation emerge from this period of economic uncertainty stronger, more competitive, and with a clear window of prosperity for all Americans.
As leaders within companies that collectively employ millions of Americans and operate and compete around the globe, we know this country can do better. Thus we offer below our ideas for (a) addressing the nation's structural challenges in a fashion that will stimulate growth and job creation and (b) stabilizing the U.S. debt:
Create Jobs and Ensure U.S. Competitiveness -- Tax Reform
First and foremost, America's corporate tax system is globally uncompetitive and is woefully out of step with the world in which we currently live. Many of the companies that are signatories to this letter and the innovations that drive their businesses simply did not exist when the code was last revised in 1986. While the code has remained largely static over time (with piecemeal patches), the environment in which our companies operate has changed dramatically.
To attract – rather than lock out – capital and create jobs here at home, we must act in our short- and long-term interest. With regard to the first, we should take immediate steps to encourage U.S. businesses to repatriate the approximate $1 trillion in accumulated foreign earnings that are locked outside of our country because of an antiquated and punitive tax code. As a nation, we are much better off with those dollars being invested here rather than elsewhere.
In the long term, we must reduce the rate, simplify the code, and strengthen incentives for job-creating activities such as R&D, and transition to a competitive territorial tax system.
Globally, in just the past four years, most of the major markets with which we compete have reduced corporate tax rates and transitioned to a competitive territorial tax system. In fact, our largest trading partners—Canada, Great Britain, and Japan—have all done so to become more competitive. Further, this exact approach to tax reform has been urged by many who have closely studied the issue including the National Commission on Fiscal Responsibility and Reform, and the "Gang of Six" U.S. Senators who proposed a solution to the debt ceiling crisis this summer.
Stimulate Growth - -Targeted Investments
Second, there are other important steps that the select committee can take to stimulate growth, including making targeted investments. Not every dollar spent has the same effect on the economy. Thus, we recommend reducing or eliminating low-impact spending in order to create the fiscal space to focus on pro-growth investments. Within our companies we continue to invest in research and development even in tough times because that investment provides a rate of return that is a multiple of the expenditure. It is important that our government does the same. Thus, we recommend greater investments in those activities with clear economic benefits because of their transformative potential. Programs that support basic scientific research, improve our infrastructure, protect our intellectual property and create a 21st century workforce are smart investments.
Finally, government spending is unsustainable. We all face difficult decisions within our companies and making cuts is often the most difficult but also the most necessary. That is the case for the U.S. Thus, as the committee works to identify the best deficit reduction approach, avoiding the hard decisions, such as entitlement reform, would be a mistake. There are steps the select committee can take that will contribute significantly to deficit reduction, and provide long-term growth opportunities for our economy. Those steps include:
- Spurring Innovation and Job Creation through Spectrum Sales: The Congressional Budget Office estimates new authority expanding the Federal Communications Commission's spectrum auction authority to conduct voluntary incentive auctions (as included in S.911) could generate more than $24.5 billion. Making more spectrum available for mobile broadband, including recognizing the benefits of unlicensed uses, is a win-win-win-win, creating jobs, benefitting consumers, fostering innovation, and reducing the deficit.
- Strengthening the U.S. workforce to raise direct revenue: Studies across the political spectrum confirm that increasing the number of employment-based visas for highly educated workers will not only strengthen our nation's workforce and competitiveness, but could also generate direct revenues for the Treasury. Simple fixes such as awarding permanent visas to foreign-born students who earn doctorates and repealing the annual limit on the number of applicants per country who receive visas is simply good economics.
- Reducing spending through information technology: The Federal Government uses information technology to great effect. As has been made clear, however, there is a gap in power and productivity between Government IT and IT across the private sector. The federal government can reduce spending by more than $1 trillion over the next 10 years by adopting the latest technology-fueled private sector best practices in key areas, such as consolidating IT infrastructure, streamlining government supply-chains, encouraging adoption of cloud computing technologies, and reducing government energy use.
As markets lose faith in America's ability to meet its challenges and lead the world, investors look elsewhere and entrepreneurs fail to dare. We can do better. We humbly offer our thoughts on a few viable, but near-term and practical solutions. More importantly, we offer our commitment to working with you to move this great nation forward towards greater prosperity."
You can read a full copy of the letter here.
The winds of change are howling. Support these initiatives - write your Congressman and tell them to buck up and take some action.
Tuesday, September 27, 2011
Kick-starting the 450mm transition and collaborating to improve the 300mm regime, the state of New York signs 5 industry heavyweights to put some big skin on the table ($4.4 billion to start). Intel is going as far to relocate their 450mm development facility to the area. Other participants include IBM, Samsung, Global Foundries and TSMC.
Governor Cuomo Announces $4.4 Billion Investment by International Technology Group Led by Intel and IBM to Develop Next Generation Computer Chip Technology in New York.
This will cause meetings.....
Thursday, September 08, 2011
Solyndra spokesman Dave Miller said the search came as a surprise, but he emphasized the company is "fully cooperating" with federal officials. He said he did not know the purpose of the search, but he speculated it could have something to do with the $535 million in loan guarantees the Department of Energy awarded to Solyndra.
It could have something to do with the $535 million? You think?
Uncle Sam... Trying to save face.
Saturday, September 03, 2011
Earlier this week Coburn Analyst Brynne Thompson penned her take-aways from a TED presentation by Tim Harford. Tim's presentation discussed the need to embrace trial and error when working in complex systems. Brynne extends the subject matter to investing (there's plenty of complexity to deal with in that arena!), highlights some of the main points of the talk and then provides a neat summary of how she plans to apply it to her work.
The first part of this is a lead-in from Pip Coburn - the main man at Coburn Ventures. His comments are always worth a read:
What follows is a note from Brynne through our friend Julien. Thank you Julien.
I am immediately reminded of a time long ago when at Lynch+Mayer the MOMENT of "taking a position" was a VERY VERY big deal that implied you had figured it out, had crossed ALL your t's and dotted ALL your I's... 15 years later I realize that I am rarely if ever in my LIFE (of which my work is a critical subset)at a point of all t's crossed and "all" my work is done AND even more so I seem to see that when I am deluded to consider that I have ALL "set" that a rude awakening is juuuuust around the corner.
Somewhere in 1996 I had a realization in doing tech - which was far more immature and far far far far far more volatile than today for others who lived it -- that when I scanned my portfolio of investments during my regular checks and thought, "No, that one is good... No need to make a follow up check," that I had somehow alerted the Gods that I didn't know my humanness and that I was due for a painful lesson... Sure enough an earnings pre-announcement would spring. (This was BEFORE the days of what I think my friend Arnie Berman dubbed as "earnings pre announcement season" - which has stuck.)
So... EVERYTHING we do today I consider work-in-process... I am today reviewing half of our short pre determined game plans to consider with Dave and Brynne especially what we might be missing...
And I love the work I get to do...
Thanks Brynne for this note... It is well-timed for me :)
And here is Brynne's note:
Subject: **Tim Harford at TED (video from Julien): The take away for me and my investing:
Develop a stronger muscle for trail and error in problem solving. Understand there is no one right answer, *especially* in complex systems.
I watched the 18min. TED talk that Julien passed on. It is on the topic of using and embracing trial and error. Fantastic video (link below).
The main point for me was that trial and error itself - or rather the thought that there can be errors at all - can be incredibly difficult, even excruciating, for most of us. And if we cannot stand to admit mistakes or defeat, we may do anything possible to avoid that experience.
I don't think I have often, if ever, heard of "trial and error" applied to investing. Investors in particular aren't fond of "error", whether tracking error or analytical error. In my mind, it's tied so closely to "poor performance!" or "lost performance!" that I want to do everything in my process to avoid error.
But, I think taking a step back from the very literal considerations around "error" can be very valuable to allow some space for all the exploration and openness that trial and error can bring to an investment process. When stripping it down, trial and error is really just iteration, and learning from each iteration, and THAT is something that we are finding to be ever more powerful, whether it is in our own investment process, or with companies that can be at "home" doing work iteratively instead of cocooning themselves off from the possibility of ever making a mistake, or companies that iterate so well that it becomes a key competitive advantage and an ability to Observe, Orient, Decide and Act (OODA) much more effectively than others.
Here are a few more points from the piece, and the video link below:
- The "God Complex": An avoidance of mistakes drives us to become "Gods" of our own little worlds - our own mental models we've created to name and explain what's our very complex world all about. When we do that, we can cocoon ourselves off, miss seeing clearly what is going on in "reality", and make poor decisions.
- Iteration/Trial and Error helps remove the "need" for a God Complex. His example is that he thinks one reason the US produces some very good businesses is that the rate of failure of US businesses is higher than the rest of the world: 10% of US businesses fail every year. There is iteration inherent in a system that allows such "failure". The turnover starts to chip away at a God Complex. There's an acceptance that starting and growing a business is tough; not everyone "wins", but also, there is no "God" example, and often, no one person in charge to say what a "successful" business is, so there is a freer definition of what business "can" be. Other business cultures around the world rely more on a few "God" examples of how to do things.
- Education: If we want to develop iterative skills and remove the cocoon of the God Complex, start teaching children in schools that there are some problems that *do not have ONE right answer*. Instead, we give exams that have question and answer with the focus on "only one right answer", that is adequate for learning some skills and memorization but inadequate for problem-solving, creativity and learning to "sit with" the uncertainty of not having an answer but being able to stick with the iteration and exploration of the problem.
- Why is the ability to use trial and error ever more important or at least, useful? The world is complex, and increasingly so! We have very interesting problems right now, and if we are so stuck and unable to admit our own infallibility, or think there is only "one right answer", we will likely be stuck with the same problems, stuck and stagnant.
TED Talks -- Tim Harford: Trial, error and the God Complex
So what does this really mean?
I loved this video, if only as a great reminder of being patient in sitting through some of the tough questions that come up for me with regard to investment process or company analysis, giving myself some space to work through the problem from different angles like a complicated puzzle instead of a short-sighted conclusion that I'm wrong if I don't "know" all the answers right away, or more likely, if I don't always see very direct, linear patterns of change, but a number of points that form a more complicated model that takes an openness to many different points of view about what a mental model is, and some injection from different mental models, disciplines and perspectives to really generate insight.
I too loved the video….
The comments about education reminded me of some of the things that were said during a panel that honored Jack Kilby and Richard Smalley at an '06 Nanotechnology conference in Dallas, Texas. After the panelists, all Nobel Laureates, described what drove them to their honored result they dove in to the topic of education. To a man they agreed that "problem solving" skills were not being developed in our educational system. This is not an exact quote but the message went something like this: "We aren't teaching our students to solve problems. We are teaching them to memorize and regurgitate." One panelist, the late Alan MacDiarmid, suggested that students should question everything - going as far to say that everything a student read from today's science text book would be wrong in 20 years.
Reader comments are more than welcome!
Wednesday, August 31, 2011
In a nutshell: Customers of Novellus have become incrementally more cautious and they have curtailed their expansion plans. CEO Rick Hill said a contagion developed amongst the customer base that was spawned by the nervousness that gripped some members of the sector in July (during Semicon West). This weakness or, fear, was described as the confidence level device makers have for end demand out nine months from now. Given the way orders are trending it is, in essence, a vote that says device makers have no confidence in the economy.
- 9.3% growth in the PC market for this year seems reasonable but those customers have been more cautious.
- DRAM pricing is weak and demand is also weak. On the other hand, logic (microprocessors) still seem reasonable because of diversity in their application.
- Not all the markets look terrible, smartphone and tablet forecasts continue to look robust.
- NAND continues to be a bright spot.
- Semiconductor utilization rates remain below seasonal levels
- Novellus expects spending for wafer fab equipment in this year to be in the range of $30 to $32 billion and expect demand to be pushed in to '12.
Guidance: For Q3 expect bookings to be down 15% to 30%. Guidance for shipments and revenues was narrowed to $300 to $320 million
EPS $.65 to .75 to the high end with the Q3 tax rate being a major influence.
This reminds me of an analogy that many industry veterans use to describe the way these cycles develop: "If I tell you it is bad and you tell someone it is bad, they tell someone else, and then that person tells someone, before you know it, it gets bad."
My $0.02 -- I don't think things are going to turn up until the early months of next year.,
During a panel at the conference Aart de Geus, chief executive of Synopsys, had this to say:
"We're not just dealing with silicon scaling complexity but with a kind of systemic complexity where being best-in-class in one area is not sufficient to avoid risk and risks are going up."
"It's a winner-takes-all situation with whole ecosystems racing to high volume systems, so value chains become very important."
A strong value chain is required today and it will be even more important in the next few years. Recall my comments the other day about the need to have everyone in sync when the industry transitions to 450mm wafers:
The chain extends well beyond the fab and the packaging house. If you are one of the remaining high volume producers of smartphones, tablets and, heaven forbid, PCs, you are bound to be a part of that chain. Hopefully your partners are all in sync.
Also on the panel was Warren East, chief executive of ARM. Supposedly he tipped his hat to Apple when he said,
"As products get more complex, no one company can provide everything, but some can provide a great deal of the system."
Hmmm... We're going to find out how good the Apple foodchain is over the next couple of quarters. Everyone knows that TSMC is the company that they want to make the next iteration of their processor (A-6). A good friend, a process expert that works right in the trench, had this to say when asked about TSMC's efforts:
I guess then the 28nm scaling of some long-in-the-tooth device technology coupled with a new form of packaging technology could provide for a perfect “s”-storm scenario. That ought to have someone at Apple concerned, I would think, not to mention TSMC. If it causes them to skip a beat on getting new products out relative to the steady drum-beat of new product announcements in the past – put in context after the Jobs resignation – and you could easily have a lot of folks shouting “See, I told ya so.” Rarely do those types do enough digging to realize the pre-cursor to such a hiccup was already set into motion while Jobs was still at the helm.
It is anticipated that the The A-6 processor, as noted on many tech sites, will hit the market in the second quarter of next year. It will be interesting to see if that time line can be met. What if it doesn't work? Will they go back to rely on Samsung?
Is the chip industry really getting close to the wall? The challenge of improving chip yields, a subject we have been discussing on my mailing list over the past two weeks, was mentioned by the panel:
"Process technology tolerances are edging closer to design rule margins, affecting chip yields which can be in single digits as new nodes first come up. That means changes in IC design can more readily impact manufacturing yields."
Single digit yields (and in some cases a bit higher) are exactly what we are hearing about today. Maybe we are getting close to the wall? Speculation that the end is near has been going on for years. Perhaps it is just going to take more time to get past the challenges?
Friday, August 26, 2011
Things I find most interesting:
- WW DRAM capacity will not exceed levels reached in ’08 (Slide 72). This says a lot about the PC business.
- WW NAND capacity will exceed DRAM capacity in Q1 ’13 (Slide 73). This speaks to the smartphone/tablet trends and the adoption of SSDs in the enterprise.
- Everything about the Embedded Solutions Group (starts on slide 25) -- Automotive!
Micron Summer Analyst Presentation (PDF Download)
Thursday, August 25, 2011
IC market to top $300 billion in 2013 - ElectroIQ.
However, IC sales will exceed $300 billion for the first time in 2013. The IC market first topped $10 billion in 1980, $100 billion in 1995, and $200 billion in 2005. While 10 years passed from $100 billion to $200 billion, the $200 billion to $300 billion jump will take only 8.
I believe Gartner (at that time they were known as Dataquest) was the first to put out a forecast that talked about $300 billion in annual sales. After the dot.com bust not many were willing to venture out on the limb.
Long range forecasting in times of uncertainty...... I dunno.... I suppose somebody's got to do it!
And the strongest growth will happen in... drumroll..... :
The largest growth for IC application segments will belong to automotive chips. Analog vehicle chips will see 32% market growth in 2011. The Automotive Special Purpose Logic/MPR segment and 32-bit MCUs won't be far behind. Sophisticated safety systems, driver information systems, and engine control units will keep the automotive IC market active through 2015
I’ve said it and heard it from others in the industry: “The Achilles heel for most capital equipment companies, and for that matter, semiconductor companies, is the inability to see changes in end demand.” Jim Bagley made this clear a few years ago at SEMI's Industry Strategy Symposium when he pretty much said he really didn’t care what was happening in the end markets because their job (at that time he was with Lam) was to stay in step with Moore’s Law.
You have to admit, they are focused.
To AMAT's credit, they have been pretty open about the weakness they were seeing in the PC markets. That really makes you wonder when Intel is going to say something negative.
On the AMAT call and the report…
I doubt that there was an investor out there that didn't expect bookings to be down and the outlook to be weak. I won't bore you with those details. I did find it interesting to hear management talk about foundries moving equipment from under-utilized trailing edge nodes, 65nm, to 28nm production lines. I’ve pinged folks in the industry to learn more about migrating old equipment to advanced lines. I know it has been going on for years but now we're talking about some serious reuse. If you have any insights please feel free to chime in.
From what I hear yields at 32nm and below are very low – some suggesting that good yields are below 20% (yes, there are some exceptions - Intel and a few of the bleeding edge memory companies but generally advanced yields are not good). I also hear from my moles that getting yields up is going to take longer than expected. KLA’s management team made this very clear on their earnings conference call. If anyone knows, they should. I posted some of KLA's comments here:
More Capital Spending Revisions - Tokyo Electron
As for next year, AMAT’s management talks about spending for 20nm and 14nm device production. Well, there are issues here – particularly with inspection and measurement – not to mention the cost of the lithography bay. A few folks in the inspection and measurement business have suggested that when the industry reaches 22nm inspection will hit the wall. We’ll likely hear more about this but I’ll say it now, “You can’t measure and you can't fix what you can’t see.”
For posterity let’s assume the up cycle kicks in again and AMAT does get orders for 20nm and below – call it a couple of years worth. After that we head to bigger, 450mm, silicon wafers. Uh, Oh! Remember this one?
When Getting Bigger Makes You Smaller
If only a few chipmakers are able to buy the 450mm tools, then the R&D costs must be passed on to that small universe. The tool vendors are bound to be impacted financially, perhaps severely. Either way, a small number of buyers means more development costs have to be absorbed by the tool vendors. Said another way - the tools are more expensive in smaller volumes.
What I am trying to get my arms around is exactly who is paying for 450mm development?
And if all the big semi-equip vendors develop 450mm tools, except for just one in the critical chain of manufacturing, then the fab can’t operate. So lithography, etch, CVD, PVD, clean, robotics must all commit to develop in parallel. And to do so they must be motivated that the timing is right and the returns are there. Further they must have sufficient balance sheet to support a long time line effort where the returns are years out.
Yes, this could go on for hours…..
Moving along, I’ve got a lot of doubts about the solar business and a big recovery. AMAT is very optimistic about solar and they believe projects will get funded. They might…. I'll admit that I am a skeptic. I hate to say it but I suspect that in just a few short years we are going to see much higher efficiencies and we will regret funding a lot of these projects. Then again, maybe they are right and the money spigots will remain open…. Regardless of the economic considerations and our fiscal issues.
Last but not least, some thoughts on the stock…. I’m not a fan. There are companies out there in chipland that are growing faster and, of all things, yielding more than AMAT so I think it is best to look at them first. Where does the stock get appealing to me? I’d like to see a 4% dividend yield – at least.
The major semi-equip investment attraction is, and has always been, the volatility and calling the turn. That’s what keeps the hedge funds engaged with the group. Without the hedgies p/e’s for the group would likely fall further. Using some of the newly revised Wall Street estimates, say $1.10 for CY11, and less than that for CY12, I believe the share price of AMAT, which is currently trading at ~9.9x that number, does not reflect larger earnings cuts ahead.
We will know when we have arrived, as the volatility will discount the p/e rather than be a premium. Semiconductor capital equipment companies are cash rich, mature and will either institute a dividend or up their dividend. As all mature companies do. Notable is the fact that 3M and CAT have higher p/e’s.
If you are interested in hearing more discussions like this feel free to join the INFRASTRUCTURE mailing list.
Monday, August 15, 2011
Excerpts are in order:
In June, the previous quarter’s report lowered the annual GaN LED revenue growth forecast to just 4% to $8.7B on rapid 1H’11 ASP reductions from a rising surplus, slower LCD and LED panel growth and the lighting market not yet cost competitive. With growth slowing, margins shrinking, the oversupply worsening and credit in China tighter than expected, LED manufacturers have pushed out a significant number of installations in 2011 resulting in a surprisingly large MOCVD shipment downgrade.
IMS Research has lowered its 2011 GaN MOCVD forecast by 24% to 833 reactors, which still represents 4% growth over 2011 as shown below. Smaller capacity growth should slow down the LED oversupply and stabilize pricing which will benefit near term LED manufacturer profitability and eventually lead to more tool sales resulting in an upgrade to the 2012 outlook.
Q2’11 GaN (gallium nitride) MOCVD shipment results whose highlights included:
- Shipments were down Y/Y for the first time since at least 2008, falling 14% if Veeco’s MaxBright reactors were excluded as Veeco has not yet recognized revenues for this new tool in according with GAAP.
- Including MaxBright reactor shipments, installations were still down 2%.
- Excluding MaxBright shipments, China continued to dominate the market, accounting for 70% of installations with Korea and Taiwan at 11% each.
- Lextar was the top customer in Q2’11 and 9 of the top 10 customers had operations in China.
- Aixtron remained #1 in GaN MOCVD shipments, gaining 4 points of market share to 57% excluding MaxBright with Veeco losing 3 points to 41%. Including MaxBright reactors, Aixtron’s advantage slips to 49% vs. 48% for Veeco, the closest it has ever been between the two competitors.
- By region excluding MaxBright, Veeco led in China, USA and Europe while Aixtron led in Korea and Taiwan. Veeco’s K465i remained the industry’s most popular tool, but Aixtron’s CRIUS II and G5 each gained share and rose to the #2 and #3 positions.
- The 4” shipment share rose more than 50% while the 6” share more than doubled as companies move to larger wafer sizes to boost their output and lower their costs.
There's even more in the full release so check it out:
IMS Research Downgrades 2011 and Upgrades 2012 Quarterly GaN LED MOCVD Forecast – New LED Entrants and 2011 Delays Boost 2012 Outlook
Aside from how this will influence Veeco and Aixtron, one should bear in mind that Applied Materials is getting in to the reactor business. Folks that are interested in how they are doing on this front should tune in to their earnings report coming August 24. Surely someone will ask questions about their progress during the call.
Not so fast!
Believe it or not, a call for this buyout was made two weeks ago. This author makes the case - pointing out many of the items being discussed today:
If you look hard enough the scoop is still out there. It's just a matter of finding it amidst all the noise.
Now you have to wonder what will happen with the chipmakers that are supporting Android? In particular, Texas Instruments, Qualcomm and Nvidia.
Thursday, August 11, 2011
Over on the EE Times website Mentor Graphics CEO Wally Rhines writes a piece about the deconsolidation in the chip industry:
Semiconductor Industry, Deconsolidation
Figure 1 shows that semiconductor revenue based upon foundry wafers grew to about 20% of total semiconductor sales by the year 2000 and then remained flat for most of the next decade. The 20% share of revenue assumes that fabless semiconductor companies sell their foundry-produced products at about a 2X multiple of their foundry cost. This 20% share has recently begun to grow again. Exceptionally large capital investments by foundries in 2010 and 2011 will inevitably increase the share of semiconductor wafers sourced from foundries.
I think it is important to recognize that this most relevant to the fabless chip houses and not those that have manufacturing facilities - foundry or IDM. Much of the consolidation in the semiconductor food chain has taken place in the capital equipment business and its supply chain.
This story was linked to another, longer article Wally released over a year ago: Is Semiconductor Industry Consolidation Inevitable?
Most interesting to me was this observation on Page 2:
Looking back across the past six decades, TI (Texas Instruments) is the only company that has remained in the top ten throughout semiconductor history. When consolidation does occur, it is mostly among those companies that fail to adapt to the next level of emerging technology.
Never under-estimate the big Texan.
While I was reviewing this I was thinking about some of the comments made by the Gartner analysts during Semicon West - see this article I wrote last month for a few of their thoughts:
When Getting Bigger Makes You Smaller
I'm torn. On the one hand I understand the opportunities the fabless semiconductor companies have and the ways they can leverage foundry technology. At the same time I believe the cost of designing a device, getting wafers processed and getting a product in to the market with enough volume to be recover these non-recurring engineering costs - much less make money - will be more challenging as the manufacturing side moves to more advanced technologies.
I also tend to agree with the Gartner folks that the cost of processing silicon at the most advanced levels will pare down the number of companies that can actually finance and build a wafer fab.
I'm open to opinions on this front....
In other stories of the week, the PC has died several times:
Mark Dean, IBM CTO for IBM Middle East and Africa writes that they are leading the way in the post PC era. Every article I've read is quoting this line:
But, while PCs will continue to be much-used devices, they’re no longer at the leading edge of computing. They’re going the way of the vacuum tube, typewriter, vinyl records, CRT and incandescent light bulbs.
Intel, obviously, takes issue with this: Intel CFO says view unshake by US debt downgrade
Families buying their first PCs in emerging markets, the source of half of Intel's sales, are unlikely to put off their purchases because of troubles in the United States, CFO Stacy Smith said in an interview with Reuters on Monday.
We'll see if demand really does hold up. I have my doubts. Several stories from Taiwan are noting weak demand (but optimism for the latter part of this year and next):
Compal sees drop in July revenues and notebook shipments
Acer unlikely to increase notebook orders until September
Are companies optimistic because they believe Ultrabooks will save the day? Intel is putting some skin on the line. Yesterday they created a $300 million development fund:
Celebrating 30 years of innovation, the PC is the ultimate Darwinian device and Intel is striving to again reinvent mobile computing,” said Mooly Eden, vice president and general manager of Intel’s PC Client Group. “In 2003, the combination of Intel’s Centrino technology with built-in WiFi, paired with Intel Capital’s $300 million in venture investments and other industry enabling efforts, ushered in the shift from desktop PCs to anytime, anywhere mobile computing. Our announcement today is about Intel mobilizing significant investments to achieve the next historic shift in computing.
There are three key phases in Intel’s strategy to accelerate its vision for this new category. The company’s efforts begin to unfold this year with Intel’s latest 2nd Generation Intel® CoreTM processors. This family of products will enable thin, light and beautiful designs that are less than 21mm (0.8 inch) thick, and at mainstream prices. Systems based on these chips will be available for the 2011 winter holiday shopping season.
One thing is for sure, Intel will not go down without a fight. A few folks believe Apple has them in their plans:
Intel will not win the tablet market with any of the various Atom chips rolling out at 32nm, 22nm and even 14nm. They are too late to a game that Apple owns 90% of today and will so in the future. All of these ultra low power atom versions are like the Saturn test rocket developments that preceded the Apollo 11 Moon Landing. They are necessary test chips where engineers at Intel try out new circuit designs and architectural tradeoffs in tuning power vs performance in preparation for a set of chips co-architected with Apple and appearing in 14nm.
Lots of speculation here...
I suppose I could post 150 links to stories that talk about the market cap gyrations between Apple and Exxon. Either that or I could tell you to calm down and quit worrying about the market action. Nah.....
Last but not least, from the humor side, a P.S to a comment on a "Buy stocks, sell bonds" post by Felix Salmon:
p.s. after the titanic hit the iceberg, on a relative basis, the stern was where you wanted to be.
Hang in there!
Friday, August 05, 2011
The release also states that Texas Instruments is the company to beat in core application processors.
Here's a link to the story and the TOC for the report:
Expectations for the 2H 2011 tablet market.
A quick search shows another story here:
Apple iPad to account for 61% share of global tablet market in 2011, says a new report from Digitimes Research
Yeah, this is going to raise some eyebrows this weekend.
1) K&S tools are shipped to assembly houses with relatively short lead times.
2) Swings in demand for wire bonding tools typically preceded up and down turns in the semiconductor cycle by a few quarters.
Pretty straightforward stuff….
As time passed we continued to add indicators to our arsenal in an attempt to improve our sense of where the industry was headed. I say "attempt" purposely here because I will readily acknowledge that no set of indicators is perfect. By no means do I want to imply that we nailed every cycle turn. That said, gathering information about run-rates and subtle nuances in the food chain became a centerpiece of our work. "Eclectic" was how some described the collection of data points we accumulated .
Up and down the food chain we searched - starting with raw materials for wafer manufacturing all the way to those OSAT houses (Outsourced Assembly and Test) that were sending board-ready devices to the subcontract assembly business. As for the eclectic, here's a few examples: We would ping companies that supplied the vegetable inkers that were used to mark bad devices during the final testing process - thinking that changes in the demand for vegetable ink (yes, vegetable ink) could be linked to variations in unit volumes. Along the same lines, monitoring die banks at the OSAT houses became important. Companies like Amkor, Silicon Precision Industries and Advanced Semiconductor Engineering often talk about die banks on their conference calls. Amkor's Ken Joyce had this to say on their July 27 call:
"All we can say is what we see, I can’t speak for the entire industry, but what we see is that there has been a buildup in the die banks in our factories, and what does that tell us, and who are the customers that are building in there? Well, they’re actually one step, once the front end makes the decision to put those into production, they take them and they give them to us and that’s good news. I think what we’re seeing there is that our customers and their customers downstream are managing their inventories really very closely, and it probably is somewhat of a reflection of what I talked about a few minutes ago, this cloud of uncertainty that’s started to rise with respect to the demand picture."
The full transcript is here:
Instead of waiting for the companies to tell you what die banks were doing we would check to see if demand was rising for wafer storage containers and die carriers. If demand was jumping there it could be construed that the OSAT houses were shelving processed wafers before they ran them through the test and assembly process - building a die bank. For finished (tested and packaged) devices, demand for die carriers would suggest there wasn't any pull through from their downstream customers.
I could go on because every layer has its own little food chain. Who supplies bearings and actuators for automation systems, metal components for process tools, lasers and alignment components for scanners, etc, etc. This is just a few of the ways to get a pulse on the chip business - that is, if you really want to dig deep.
Most folks have no inclination to spend their days digging around in the semiconductor food chain. It's a lot of work. If you choose that route but still want to have a good idea of what is happening in semiconductor land then I would suggest you tune your attention to Microchip Technology. Microchip has proven to be accurate at calling cycles time and time again. Their management team, in my opinion, is the "Canary in the Coal Mine."
If you would have listened to Microchip when they pre-announced July 11 you probably would have sold every chip-related issue in your portfolio (and been a lot richer today). Last night the company issued their quarterly report and here's what CEO Steve Sanghi had to say:
"We believe that our June quarter results reflect weak overall market conditions, which we believe will impact the broad-based semiconductor industry in the June or September quarter, depending on the individual market exposures and revenue recognition practices of the company. Since then, we have seen this sentiment reflect to the earnings season with a weak guidance coming from Linear, Freescale, Texas Instruments, Silicon Labs, STMicroelectronics, Intersil, Micrel, NXP Semiconductors, IDT, Power Integration and others. The semiconductor supply chain is confirming the same sentiment with weak guidance from TSMC and UMC on the wafer foundry side, Teradyne from the test equipment side, [indiscernible] in Applied Materials from the fab equipment side and Arrow Electronics from the component distribution."
"There could be some companies that are exceptions due to a particular product transition or momentum in a given application segment, but we continue to believe that most companies will see the weakness manifest itself as the quarter progresses and their guidance is likely to be too aggressive."
"The broad-based global weakness seen [ph] has also been reflected by large multinational industrial conglomerates like Emerson Electric Co. Last week's Durable Goods report also cited that new orders for durable goods fell unexpectedly in June and a weak June quarter GDP report and a very weak ISM Index this week caps it all."
"Our long-term investors and analysts have seen Microchip experience these industry events first. Microchip tends to see changes in business conditions earlier than most of our peers due to a number of factors. Number one, we recognize distribution revenue on a sell-through basis worldwide. Number two, we run very short lead times, which gives customers more time to make purchase decisions based on more current business conditions. And number three, we have a large number of small- and medium-sized customers who tend to be quicker in adjusting their inventories."
The full transcript is essential reading for anyone interested in the sector. You can find it here:
I would also suggest that you put the shares of Microchip on your radar screen. In fact, run some comparisons against the other household names in the semiconductor business. Microchip has a stellar track record of profitability, stock performance and a 4.40%+ dividend to boot.
Full Disclosure: I have no position in Microchip or any of the companies mentioned in this note and I won't have any positions until I hear the Canaries singing a different tune.
Thursday, August 04, 2011
"But I have some concerns about the growth next year because capital investment in foundries in 2010-11 has been very high and there is some worry that there would be some oversupply."
He's worried about a price recession for devices in the second half of next year although he expects that dip to be overcome by demand growth in those markets we hear about so often - mobile, automobile, and security - to name just a few.
He goes on:
"The problem is that you have doubled the investment in foundry wafers while investment in memory is very modest. So I would expect the fabless companies will find that there is a plenty of capacity starting second half of next year leading a likelihood of price competition in fabless semiconductor industry."
Putting further spin on the situation he notes that this would be short-lived as long as we have economic growth. Well, that's certainly true. If the economy grows folks spend money on electronic gadgets. Growth in the semiconductor industry is totally linked to growth in worldwide GDP - an irrefutable fact.
I've written recently that during the first half of this year the foundries have dramatically slowed their rate of capital spending. Aside from the ramp that is taking place at Global Foundries there just hasn't been a lot of buying. Last night we heard from United Microelectronics Corporation - presentation materials from the report are available here and here.
In a nutshell, flat sales, declining profits and factory utilization rates of less than 75% are what to expect for the next three months. No predictions were made about the fourth quarter - which is similar to what we’ve been hearing from everyone else in the industry.
As for their capital spending plan, UMC is sticking to their $1.8 billion target for this year as they ramp their 40nm process technology. They will also begin to install 28nm production tools late this year for production ramps next year.
The concern I have with overcapacity rests on two items: 1) We need healthy Global GDP Growth to absorb the production and 2) How improving foundry yields will impact the number of devices generated at the 40nm and 28nm nodes. The latter point being one I mentioned in a previous note. These yields are bound to go up as companies move along the learning curve.
Circling back, as the market swan dives I'm getting more interested in the EDA companies. Mentor Graphics, Synopsys, Cadence Design Systems and Magma Design Automation are on my radar screen. Shares of the largest player in this field, Synopsys, are really starting to get interesting. On August 17 the company will provide an update. I will be most interested in the forward view - assuming the world hasn't ended by then. :-)
"As shown in Figure 1, Intel remained firmly in control of the number one spot in the ranking. In fact, Intel extended its lead over second-ranked Samsung by registering a 43% higher sales level than Samsung in 1H11 as compared to a 24% margin for all of 2010."
The rest of the story is here:
First up, another story about OMAP from Texas Instruments heading to Google's Ice Cream Sandwich and possibly even the Nexus:
Google Chooses TI's New OMAP to Lead the Ice Cream Sandwich Charge
Next, a slick graphic from Wordstream showing the Top 20 Most Expensive Keyword Categories in Google AdWords:
Most expensive: Insurance, Loans, Mortgages, Attorneys and Credit....
Wednesday, August 03, 2011
From the release:
“In addition to uncertainty of global economy, demand for consumer electronics such as smartphones and tablets is slowing down compared to our initial expectations at the beginning of the year. Hence the semiconductor price is falling and fab capacity utilization is declining, we expect it to take some time for the SPE market to recover.
With these trends, we are currently assuming that the scale of SPE capex this fiscal year will decline by around 10% in contrast to our initial expectation of a 10% increase.
As we enter 2012, we expect a number of factors to stimulate chip demand including Windows8, further evolution of new mobile devices such as iPhone5 and new tablets, as well as the 2012 Olympics and the US presidential election."
More detail can be found in the Management presentation here:
The bad news about capital-spending is piling up and the stocks remain under pressure. The Applied Materials report will be the icing on the cake. After that, investors should be looking for the bottom in share prices.
Don't touch that dial!
Saturday, July 30, 2011
We are just starting to hit the ramp phase for the Tablet business. Today we hear from Cisco and their "Cius" Android Tablet. Yes, Cisco is getting in the game with everyone else.
The holiday selling season is right around the corner and all those products that were announced earlier this year are about to hit the shelves of a store near you.
Here are a few more notable points from the press release:
- Total Tablet PC panel production rose to 6.43 million units - up 28% month-to-month.
- 5.38 million 9.7-inch panel displays - the size used in the iPad - increased 26% month-to-month.
- 10.1-inch panel shipments recorded 986K units, showing 37% month-to-month increase. The panels will be used in Samsung's Galaxy Tab 10.1.
And now, a bit of eye-candy (click image to enlarge):
The Source: 'Monthly Large-area TFT-LCD Panel Shipment Data Report',
Is it really a surprise? 9.7 inch panel production is 84% of the market! That really speaks volumes (pun intended) about the dominance of the iPad.
Friday, July 29, 2011
Many of the company presentations contain the same message - tough sledding is expected for the next few quarters. But never fear, you will also find optimistic expressions about the long term view.
Taiwan Semiconductor Presents
Here's a Global Foundries Video - (friends tell me they love marketing). And while we're at it, here's a presentation from Semicon West on 450mm - apparently they've moved it up on the priority list.
Speaking of the foundries - here's a roundtable discussion about multi-foundry strategies.
Lam Research Quarterly Report Presentation and their Analyst Meeting Presentation given at Semicon West.
I tuned in to the Lam conference call. The comments that stuck out for me centered on lead times. Here's a snippet:
"In addition, given the relatively short lead times for most equipment, customers have the ability to better time their deliveries to match the expected demand environment. As a result, even with overall healthy wafer fab spend levels, we expect a quarterly shipment variations in the range of plus or minus 25% will be experienced. "
Excerpt from Seeking Alpha
Shipment variations like this will drive those estimating quarterly earnings nuts and, to some extent, rendering them meaningless. Investors will have to finesse their trades - stepping in to buy when things look terrible and selling when things look great. It's the nature of the beast…..
KLA-Tencor was out last night - hear that echo? Slowing….
Think back to earlier in the year, during CES and Mobile Congress. Every company was blowing hot air about the 28nm ARM processors they were going to deliver for this year's holiday selling season. Well, 28nm has been a big challenge for the foundries and the high volume production ramp for many of those designs has been pushed out in to next year. An excerpt from the KLA conference call is in order:
Unknown Analyst -
Okay. One last question. On the 28-nanometer ramp, compared to the 40 nanometer, we know that foundries have suffered quite a bit on the yield issue in 40 nanometers. Based on what you've seen on 28, is the yield issues getting more difficult, or it's pretty much at the same level?
Richard Wallace (KLA CEO)
When I talk to customers -- and we have a lot of conversations the last couple of weeks, I'd say more difficult, more challenging. And 40 was interesting because as you know, a lot of people are going to do 45 and then they move to 40, and they were surprised by how difficult the yield was. And that was the catalyst for a lot of business for us. But if you look at 28, it's actually harder. Now they're a little more aware of the challenges, but we're getting pushed very hard to bring out our latest capability and technology to get it in line because I think our customers are pretty aware that they're missing critical defects in areas that are going to cause problems for them. And there's also a metrology challenges. So I think more difficult and a lot of interest in technology buys. And obviously, if we did have a prolonged softening, I think the technology position for K-T plays pretty well into that space because of the yield challenges they're seeing. No question.
Excerpt again from Seeking Alpha
Forecasting firm IC Insights has lowered their estimates for this year's Worldwide Semiconductor Market :
"Although the U.S. debt ceiling and European debt crises are likely to be addressed without creating a worldwide financial panic or meltdown, both of these situations have caused a great amount of uncertainty throughout the world, affecting both businesses and consumers alike. The problem with uncertainty in the marketplace is that it typically results in "hesitation" and "worry" for consumers and businesses. This environment of hesitation and worry is not conducive to good economic growth and can sometimes have a worse impact than if the negative event itself took place!"
A picture from the release:
All in all, a tough week - and I didn't even bother to touch on the mess in Washington. Keep the faith though. It will get better. Technology and chip manufacturing will continue to roll forward - with a few bumps along the way. Today's sentiment clearly lands on the bearish side of the ledger. That's a good thing if you are looking to pick up some bargains.
Wednesday, July 27, 2011
Still digging through the stories from the last few weeks. With Lam Research reporting this evening it's worth reviewing what Applied Materials had to say at Semicon West. Lam, no doubt, will echo some of these sentiments.
- Eight new products
- Soft Industry Outlook in the Near Term
- Long term Applied is bullish (no surprise here)
- Over $100 million to be spent on 450mm development in 2012
- Reiterated that 450mm will happen. (This week I've been talking with several industry folks that work with Intel and they are saying the 450mm push is on - I'll come back to these discussions in another note)
- Foundries have fab shells ready but are in a lull. Applied did say they could come back quickly though. (My $0.02 - The foundries have not been spending much money- in fact, by some estimates foundry orders in the 2nd quarter 2011 were at their lowest level since the third quarter of 2007 and the 1st quarter of 2006 - this does exclude the horrific low made in 2009)
- DRAM companies are also reluctant to spend - noting weakness in the consumer PC market.
I've been expecting a year end bust-the-capital-spending-budget but given the questionable macroeconomic environment I'm beginning to wonder if that is going to happen. Granted, not everyone is holding back. Intel is still investing heavily - which is really not unusual because they typically invest counter-cyclically. Some folks are even going as far to suggest that Intel is on allocation.
Many of the reports and comments I have received are suggesting that the next upleg for capital spending will be pushed out in to the early months of next year. That's not that far away when you think about it. I say this because the stocks tend to start moving before the cycle kicks in. If the script plays out like it has in the past then it might be a good idea to put these on your radar screen. When confessionals are over - Applied reports in August - we might see some bargains out there.
Tuesday, July 26, 2011
SEMICON West - TechXPOT North Two: Test in Transition: Changing Customer Requirements.
Several comments from the presentation made by Dale Omhart (Texas Instruments) really stood out:
Test contribution to total semiconductor manufacturing costs continues to INCREASE
Memory Test cost may have decreased by 50% in the last 10 years
In most cases, Test does not add value to the finished chip:
- This is especially true in NonMemory products
- Therefore, any test cost is painful to chip makers
Semiconductors are becoming more customized:
- There will be a very few high‐volume applications
- Big consumer products like iPhone (being an exception)
This evening ATE provider Teradyne will provide investors with an update of their business. The Teradyne report follows right on the heels of the quarterly release from Advantest. If you will recall, Advantest recently closed the $1.1 billion acquistion of Verigy. Their report was solid but the outlook, like so many in the semiconductor food chain, they expressed caution about the future. Orders are expected to decline and sales will be flat. The report, with notes, is here:
Advantest Presentation with Notes
From slide 12:
"...semiconductor manufacturers are shifting to a wait-and-see stance on capital investment and utilization rates are falling, in step with deteriorating end-user demand for PCs, smartphones and other products."
"In non-memory testers, the company expects demand will remain high, led by the MPU test segment. However, softer demand is foreseen in the communications and consumer electronics device test segment."
This last bullet is telling. The MPU market is where you find Intel and the folks that are cracking out ARM processors. It's also a place where Teradyne is very strong and I am sure they will say something about this in their report and on the conference call.
Thursday, July 21, 2011
Well, that analogy held true over the years. In the mid-to-late 90's the rising stakes flushed the Japanese right out of the DRAM business. Sure, Toshiba is in the memory game but at that time I was focused on DRAM. NAND Flash was just a blip on the radar screen.
Does the analogy hold true today?
After Intel said they were raising their capital spending budget (again) I thought a bit of a review of some of the predictions from last week would be appropriate. Yesterday I pointed to a 450mm panel discussion at the MCA BrightSpots Forum. Today, I've dredged up some comments from the folks at Gartner.
At the 450mm Panel Session Bob Johnson (no relation) had this to say:
- Equipment for each node costs more than previous nodes
- Periodic wafer size increases needed to compensate for more expensive nodes
- Can save 30% on production costs (target)
450mm has nothing to do with Moore’s Law. It's all about economics.
Later in his presentation he mentions the narrowing number of players:
Less than 10 companies are currently building leading edge fabs:
- 1-2 Logic IDMS (Intel and Samsung)
- 2-3 Foundries (TSMC, Globalfoundries, plus UMC or SMIC)
- 4-5 Memory Companies (Samsung, Hynix, Toshiba (w. Sandisk), Micron/Nanya/Inotera, Elpida/Rexchip
The semiconductor equipment companies say they will be ready when customers decide to make the move. I am sure they will - after a lot of kicking and screaming. The development costs are going to be enormous. Several suppliers that I talk with on a regular basis are now saying that 450mm has been moved up on their priority list.
At another session called "Contemporary Packaging: Challenges and Solutions for 40nm and Beyond", Gartner's Jim Walker served up some more food for thought:
The Bottom Line on Design Complexity is:
- Increased costs combined with reduced growth will cause significant consolidation in chip vendors…
- The industry will lose nearly one third of all semiconductor vendors by 2020
Wow! One-third going out of business? Well, that must all be due to, yes, you got it:
The Increasing Scale of Manufacturing
Here is the bottom line:
- Between the 45nm and 8nm nodes, logic fab costs will double to $10 billion. But a 31x decrease in cost per transistor will be maintained.
- Only four companies will be able to follow Moore’s law by 2018
- The annual number of new fabs built will fall by 60% between 2011 and 2015
- By 2015 foundries will account for about one third of the value of all semiconductors compared with about one quarter today
- By 2012, over 50% of packaging/test (SATS) will be outsourced
- By 2015 collaborative R&D will save the semiconductor industry over $30 billion in annual R&D expenditure
Well, ain't that just peachy? We'll be down to four companies that can pursue Moore's Law by '18? So, all the development work for advanced lithography (EUV) - in addition to the 450mm transition - should really come out of 4 buckets. I mean, who else other than Intel, Samsung, TSMC and perhaps Global Foundries, has the skin to step up and play that game?
I can see the number of new fabs falling off - I have no issue with that. Why build a new fab when you can buy an old one? Old fabs and fab tools don't die - they just get refurbished and reused. It doesn't just happen with really old stuff, it even happens with 300mm wafer process equipment. Texas Instruments is fine example - they are cracking out analog parts with tools from the first 300mm DRAM fab (the fab was once owned by Infineon and then eventually spun out as Qimonda - TI bought the equipment for a fraction of what it originally cost in a firesale).
I have a lot of questions about these predictions and I am definitely open to hearing what others think.
Feel free to chime in!
Nokia Rises, Q2 EPS Beats, Steps Up Cost Cuts
My $0.02: The bar of expectations had been lowered, perhaps even buried, to a point where it was impossible to trip over.
IHS iSuppli talks DRAM pricing and capex:
“In the wake of forcefully pursuing lithography reductions in late 2010 and early 2011, the DRAM industry is expected to employ a less aggressive approach to lithography migration throughout the rest of the year,” said Dee Nguyen, memory analyst at IHS. “DRAM capital expenditures are expected decline by 30 percent in 2011 compared to 2010. As a result, the rate of DRAM cost reductions also will slow during the remainder of 2011 and 2012. However, IHS expects that DRAM cost reductions will speed up again in 2013 as lithography shrinks return, due to increased capital spending. Spending will increase by 23 percent in 2012, which may spur steeper price reductions in 2013.
Delays in DRAM shrinks have ramifications for ASM Lithography, Tokyo Electron, Applied Materials, Lam Research and others in the semiconductor capital equipment business. Thank goodness Intel is bellying up to spend some money.
Before leaving this, one more thought. I've been talking about the "quarterly" focus and how it can cause you to lose sight of what is really taking place. Here's another quote from the IHS iSuppli release:
"Surprisingly, DRAM suppliers remained profitable in the first quarter, boosted by the substantial margin built up during the previous up cycle for the market, which began in the second quarter of 2009 and lasted until the end of the rally a year later."
Right on queue, two Taiwanese DRAM houses announce first HALF losses and capital spending reductions:
Nanya Technology suffers 1H11 net loss of nearly NT$17 billion
"Nanya expects a capex of NT$12 billion in 2011, shrinking 47.8% form 2010. With 42nm being mainstream technology currently, Nanya brought 32nm process into trial production in July 2011."
And this one as well:
Inotera Memories sees 1H11 net loss of nearly NT$8 billion
"Inotera expects a capex of NT$17 billion in 2011, shrinking 69.1% form 2010. The proportion of total input of 12-inch wafers for 42nm technology increased to 50% in June 2011, Inotera said. Inotera said it will start trial production based on 30nm technology in the third quarter and kick off volume production in 2012."
Microsemi Makes Hostile Bid to Buy Zarlink
There will be more deals. Bet on it.
Yesterday, from the mighty metropolis of Danbury, Connecticut, semiconductor material supplier ATMI hit the wires. CEO Doug Neugold had this to say about the second half of the year:
"However, we agree with reports that suggest there may be incremental softening in overall wafer starts in the second half of 2011. ATMI's focus on leading-edge solutions should help minimize the impact of any industry softness in the second half."
The full story is here: ATMI Reports 2nd Quarter and First Half 2011 Financial Results
ATMI is a nice company. Keep an eye on it.
And another nice company from the North East:
MKS Instruments comments on the second half for their semiconductor business:
"After a robust first half of the year, we are seeing some softening in the semiconductor market as the record shipments of the last few quarters are assimilated and brought on line."
The morning stock chart for MKS is a total hoot:
To round out the day, something from the wildlife, er, Solar, front:
4 Major U.S. (Solar) Projects Get Approval
"U.S. Secretary of the Interior Ken Salazar recently announced the approval of four new projects on public lands, the launch of environmental reviews on three others, and the next step in a comprehensive environmental analysis to identify 'solar energy zones' on public lands in six western states."
Animals that have a vested interest in these projects include: Mohave Ground Squirrels, the Desert Tortoise, Burrowing Owls, Flat-tailed Horned Lizards, Sage Grouse, Golden Eagles, Bats, Kangaroo Rats, Milk-Vetch, Fringe Toads and Horn-Tailed Lizards.
There will be more.....